The proposed research is concerned with the effects of extrinsic rewards, and other forms of external controls, on subsequent intrinsic motivation, assessed in later situations in the absence of salient external rewards or contingencies. The research attempts to determine the conditions under which extrinsic rewards will enhance and the conditions under which estrinsic rewards will reduce subsequent intrinsic interest in the activity on which these rewards were previously contingent. In particular, the research seeks to investigate the "overjustification" hypothesis, generated by a self-perception analyses of tangible rewards, which suggests that the provision of extrinsic rewards for engaging in a task of initial interest may actually undermine later intrinsic interest in that activity. In one series of studies, the long-term effects of experimental variations in rewards, during individual experimental sessions, are examined through unobtrusive measures of children's free-choice behavior in their regular classrooms two weeks after these individual sessions. In other studies, the effects of the introduction and subsequent withdrawal of extrinsic reward programs within ongoing classrooms are assessed relative to pre-experimental baseline measures of interest in normal classroom activities and relative to appropriate control groups. The proposed studies emphasize both the theoretical consequences and the practical implications of this line of research for our understanding of the long-term effects of rewards on human behavior.